A Randomized Controlled Trial of an Information Prescription for Pediatric Patient Education on the Internet FREE

Information therapy was first described in 1992,¹ but was formally defined in the literature in 1994 as "the therapeutic provision of information to people for the amelioration of physical and mental health and well-being."² Its aim is " . . . to increase the knowledge of the general public and to create a sense of responsibility in patients for preserving and maintaining their own wellness."² Whereas information therapy looks to improve the health of a community, an information prescription (IP) looks to improve a specific individual's health. An IP is the prescription of focused, evidence-based information to a patient at the right time to manage a health problem.³ An IP should be decision focused, evidence based, reviewed by experts, referenced, current, free from commercial bias, and user friendly.⁴ The National Library of Medicine has described their MEDLINEplus service as being "Your Information Prescription,"⁵ and has recently begun a "health information prescription" program for internists.^6- 7

The literature shows that well-informed patients have better outcomes than those who are not well informed.^8- 9 One study describes a consultation service where health care providers write an IP and medical librarians "fill" the prescription with patient-specific authoritative information.¹⁰

The Internet is now a mainstream information tool, as 60% of Americans have Internet access and 67% expect they can find reliable health information on the Internet. The next time Americans have a health question, 31% will go online first to seek an answer.¹¹ Parents (70%) use the Internet more than nonparents (53%). Parents represent almost 45 million online users in the United States today, and 19% say it improves the way they care for their children's health.¹² In a study of adult patients in a primary care general medicine practice, 53.5% used the Internet for medical information on a broad range of topics. The authors concluded that, " . . . providers . . . should be prepared to offer suggestions for Web-based health resources . . . ."¹³

We conducted a MEDLINE search on March 26, 2003, using the search terms information therapy and information prescription, and found that no studies to date have evaluated the potential changes in attitudes and behaviors of parents of children seen in general pediatric practices who are given an IP by their physician. Therefore, the goal of this study was to determine whether an IP given by a primary care pediatrician in a general pediatric clinic would change the attitudes and/or behaviors of parents of these children about using Internet health information resources.

This study was conducted in the general pediatrics clinic of a Midwest academic medical center that experiences 18 000 visits per year and serves the general pediatric needs of children in Iowa and parts of Illinois, Minnesota, and Missouri. Of the clinic population, 94% has some type of health insurance, and 6% consists of self-paying patients. Patients are white (82.2%), African American (3.2%), Hispanic (2.7%), Asian (2.1%), and of other race/ethinicity (9.8%).

All parents of patients at the general pediatric clinics of The University of Iowa (Iowa City) faculty were approached after the check-in process by a trained research assistant (S.L.K.) from September 9 through November 1, 2002. The study was briefly explained, and if the parents were interested in participating, the full study was explained and informed consent was obtained. Only 1 parent was enrolled if 2 or more were present. A parent for this study was defined as a parent or legal guardian. Enrollment criteria included willingness to enroll, residence with the child, English language (the investigators were not fluent in other languages), a working telephone, and availability 2 to 3 weeks later for a telephone follow-up survey. The study was approved by the institutional review board, and all surveys and procedures were pilot tested.

A randomized controlled trial was conducted without blinding. Control and intervention study forms were placed into separate envelopes that were then shuffled and placed into a box. Sequential packets were then drawn after informed consent was obtained (Figure 1).

The control and intervention groups received a preintervention paper survey that contained 19 items and took approximately 4 to 5 minutes to complete. It contained demographic information, asked questions related to Internet and computer use, and asked about previous use of IPs and a pediatric digital library (described in the following paragraph).

The intervention group was offered hands-on training on using the Internet to search for answers to child health questions using GeneralPediatrics.com, a pediatric digital library (available at: http://www.generalpediatrics.com) as an example resource. This was performed in the patient's room or another private room next to the clinic after the IP was given. The training took approximately 7 minutes and included an overview of the contents and navigation of the digital library. Parents could ask questions at any time. All intervention parents also received a handout containing a brief written description of the computer training and a list of community locations for computer access such as public libraries and schools.

During the normal patient encounter, 10 general pediatric faculty members practicing in their own clinics gave the intervention group an IP that was attached to the clinical chart. The IP contained 3 preprinted Web sites where patients could find authoritative patient education information on the Internet, including http://www.aap.org (by the American Academy of Pediatrics), http://www.generalpediatrics.com (by GeneralPediatrics.com), and http://www.medlineplus.gov (by MEDLINEplus). These Web sites have consistent quality standards that include authorship, sponsorship, currency, and attribution. The pediatricians were encouraged to add other Web sites or to cross out those they did not want to use. The IP was provided as a supplement to any other verbal or printed information provided to the patient and parents during the clinical encounter, and was presented as a resource to find more information about specific questions or problems including issues of well-child care. The study had been discussed with the pediatricians, and they were offered computer training before initiating enrollment. This training was similar to the parents' training but with an emphasis on Internet patient education information.

Both control and intervention groups were contacted by telephone approximately 2 to 3 weeks after their clinic visit. If it was an inconvenient time, the research assistant recontacted the family at a time suggested by the family. The telephone survey contained 12 questions for the control group and 19 questions for the intervention group. Completing the survey required 5 to 8 minutes. Some questions were scored on a Likert scale from 1 to 4 or 1 to 5, with 1 being negative and 4 or 5 being positive, and are reported as mean scores.

We coded the data using a predetermined coding schema. Data were abstracted into a coding form and entered into a Microsoft Excel database (Microsoft Corporation, Redmond, Wash). We conducted data analysis using SAS Statistical Software (SAS Institute Inc, Cary, NC) and included the Fisher exact test, χ² test, Mann-Whitney test, Wilcoxon signed rank test for paired samples, and unpaired t tests.

Of the telephone questions, the main outcomes of interest were the following: (1) Did the intervention group use the Internet for general and child health information more during follow-up than the control group? (2) Did the intervention group use the recommended Web sites on the IP? and (3) Did the intervention group perceive the IP to be valuable enough to plan to request an IP in the future, and/or recommend others to request an IP from their own health care provider?

Of the total of 250 families who were approached, 238 were eligible and 208 (83.2%) enrolled (Figure 1). Eleven parents dropped out after enrollment, 5 from the control group and 6 from the intervention group, because they did not return the preintervention survey or could not be contacted by telephone despite multiple attempts. The final groups consisted of 100 control parents and 97 intervention parents.

The demographics of the control and intervention study groups are shown in Table 1 and are similar to the overall clinic population and that of a previous study of the clinic population conducted by the authors (data not shown). The control and intervention groups were not significantly different on any demographic variables measured (Table 1). Of the 36 preintervention survey variables describing information-seeking behavior (Table 2), 3 reached statistical significance. Using the .05 level for α and random assignment to groups, one would expect at least 2 of every 40 independent comparisons to be significantly different. Health care providers and the Internet were the most common and preferred sources of health information. Home, work, and libraries were the most common places to have Internet access, and no parent reported having no access to a computer. Overall, 134 (68.0%) had used the Internet in the past 6 months for any health information and 104 (52.8%) had used it for children's health information. The last time parents went online regarding children's health or advice, they were most commonly looking for information regarding specific illnesses or conditions, child growth and development, or parenting. Most parents felt that having an IP would be helpful. At follow-up, Likert scale scores showed that parents in both groups were similarly concerned about their child's health (intervention group, 2.43; control group, 2.11; P = .16) and were happy with the information they had received during their child's visit (intervention group, 4.71; control group, 4.62; P = .27).

For the intervention group, the pediatricians recommended a total of 10 unique Web sites on the IPs. The preprinted Web sites accounted for 218 (86.9%) of the 251 Web sites prescribed (ie, a pediatrician could prescribe multiple Web sites per IP). At follow-up, the intervention group reported using 20 unique Web sites and search engines a total of 77 times. They used the IP Web sites recommended by the pediatrician for 51 (66.2%) of the 77 total uses. The preprinted Web sites accounted for 49 (96.1%) of the 51 IP uses. In contrast, the control group at follow-up reported using 23 unique Web sites and search engines a total of 28 times. The preprinted Web sites accounted for only 1 (3.6%) of the 28 total uses. The intervention group used the Internet more than the control group for general health information (44.3% vs 33.0%; P = .05) and child health information (39.2% vs 14.0%; P<.001). All parents were asked before the intervention to report how helpful they thought the IP would be. The intervention parents were again asked how helpful the IP was after the intervention. The intervention group was slightly more positive in their responses after the intervention than before, but this difference did not reach statistical significance on the Likert scale (preintervention, 3.09; postintervention, 3.94; P = .31).

During the follow-up period, only 31 (32%) of the intervention parents reported using the IP (Figure 1). Bivariate comparisons between the IP users and nonusers (subgroups of the intervention group [Figure 1]) for each of the demographic and presurvey variables failed to detect significant differences that might predict use of the IP during the study (data not shown). At follow-up, 27 (87.1%) of the IP users specified by name at least 1 of the Web sites that had been prescribed by their pediatrician. The IP users went to the Internet more frequently than IP nonusers for general health information (97% [n = 30] vs 17% [n = 11]; P<.001) and child health information 97% [n = 30] vs 9% [n = 6]; P<.001). The IP users were more likely than the IP nonusers to state that they would use the IP again in the future (Likert scale score, 4.19 vs 3.57; P = .02). The IP users also had already recommended the IP resources to family or friends during the follow-up period more than IP nonusers had (29% [n = 9] vs 3% [n = 2]; P = .001). However, IP users were as likely as IP nonusers to ask their own health care provider for an IP (Likert scale score, IP users vs nonusers, 2.43 vs 2.53; P = .66). The IP users were also just as likely as IP nonusers to recommend to their family and friends to ask for an IP in the future (Likert scale score, IP users vs nonusers, 2.51 vs 3.02; P = .41). Within the subgroup of IP users, preintervention and postintervention paired comparisons showed that IP users rated the helpfulness of the IP significantly higher on the postintervention question survey (Likert scale score, preintervention survey, 3.19; postintervention survey, 3.93; n = 31; t = 3.14; P = .004).

Although they had not used the IP during the follow-up period, some of the IP nonusers thought they would still use the IP in the future (n = 41 future IP users [64%]) and others thought they would not (n = 23 future IP nonusers [36%]) (Figure 1). Future IP users were more likely than the future IP nonusers to have already recommended to family or friends that they ask their health care provider for an IP (12% [n = 5] vs 3% [n = 1]; P = .01) or to plan to do so in the future (Likert scale score, (2.51 vs 3.02; P = .05).

The goal of this study was to determine whether an IP given by a primary care pediatrician in a general pediatric clinic would change the attitudes and/or behaviors of parents of pediatric patients about using Internet health information resources. This is the first reported randomized controlled trial of an IP in a primary care clinic, to our knowledge.

For parents of the pediatric patients in this study, health care providers and the Internet were the first and second most common and preferred sources of health information, a finding consistent with those of a study of adult patients in a primary care internal medicine practice and national surveys.^11- 13

The IPs in this study are associated with specific parental attitude and behavior changes resulting in increased Internet utilization for general and child health and utilization of specific prescribed information resources. The intervention group reported using the Internet for general and child health information and using the Internet more than the control group during follow-up. Also, 66.2% of the total Web sites used by the intervention parents were recommended by the pediatrician on the IP. As parents' perceptions of the value of a particular information resource were not explored, we can demonstrate a health care process improvement by a pediatrician-given IP in this study, but we cannot demonstrate a direct health benefit to the parent or the child.

Only 33% of the intervention group had actually used the IP during follow-up. The IP users seem to believe the IP is valuable because they used the Internet more than IP nonusers for general and child health information, they stated that they would use the IP again in the future, and they had already recommended it to family and friends. The IP users were not more likely to ask their own health care provider for another IP, indicating limited perceived value or possibly unwillingness to ask for this information directly from their health care provider. This suggests that some parents will use the IP if provided, but they will not specifically request one. Unfortunately, none of the preintervention survey variables predicted IP use in this study sample, and thus we cannot define a parent group in which the IP is more likely to succeed as an educational intervention. It is uncertain whether these findings generalize to populations displaying more heterogeneity in the demographic characteristics tracked in this study. This study included mainly well-child care visits, and, therefore, parents may not have been as concerned or had a specific, imperative, "need-to-know" question. We speculate that IPs may be more helpful for parents whose child has a new diagnosis of a specific health problem because they may be more motivated to use the IP. Presurvey data are consistent with this because almost half of the parents looked for information about a specific illness or condition the last time they used the Internet for children's health information. In addition, of the parents who had dealt with their own or a family member's serious illness in the past 2 years, 63% to 65% said the Internet played a role in their coping.¹²

Educating families to evaluate the quality of Internet health information effectively is challenging. Many criteria can be used, but most include authorship of the information, sponsorship of the Web site, currency of the information, and attribution of facts through references. Even these simple criteria can sometimes be difficult to assess on a Web page, especially because most people find information through a search engine and are taken directly to an inner page of the Web site, thus bypassing the home page.¹⁴ Nonetheless, numerous patient education handouts teach patients how to evaluate Internet health information.^15- 19

This study confirms that parents are searching for general and children's health information and that parents like the health information from their health care provider best. Giving an IP to parents certainly does not guarantee that inferior-quality information is not read or even used. However, our results show that IPs appear to influence parental behaviors regarding Internet use, and therefore pediatricians can help to guide parents to high-quality Internet resources. Pediatricians should be prepared to ask questions about Internet use and offer guidance. We believe that IPs can be effectively implemented in a pediatric practice; pediatricians can not only write but also fill their own patient IPs. Two IPs are created on a single sheet of paper, making IPs inexpensive. The IPs used in our study were attached to the medical chart but could also be available in patient rooms. An example of the IP currently implemented in one of our practices (D.M.D.) is available from the Internet.²⁰ On one side of the preprinted IP is a list of the common Web sites of the pediatrician's choosing, with the opportunity to add other resources tailored to individual patient needs. On the other side is information on evaluating Internet health information and the locations of computers available for community use. Training parents in an office setting may not be feasible, but locations offering community computer use frequently offer training.

The limitations of this study include that it was performed in an academic faculty practice that has a patient population that is higher than the national average on socioeconomic and educational variables. The study was performed with English-speaking parents only, and most visits were health maintenance visits; therefore, the results may be different in non–English-speaking populations and for acute care visits. Although intervention parents may have felt some social acceptability bias toward answering that they had used the Internet for child health information, in subsequent questioning, almost all of the intervention parents and all of the control parents could name specific child health Internet Web sites they had used. Parents' actual Internet use could not be verified, and only short-term follow-up measures were collected.

Parents in pediatric practices commonly use the Internet for general and children's health information. Information prescriptions are associated with specific parental attitude and behavior changes, resulting in increased Internet utilization for general and child health information and for specific high-quality information resources. Pediatricians should be prepared to ask questions about Internet use and offer guidance to high-quality Internet health information resources using IPs or other educational interventions.

The Internet is now a mainstream information tool. In this study, IPs were associated with specific parental attitude and behavior changes resulting in increased Internet utilization of high-quality pediatric information resources on the Internet. We suggest a way for pediatricians to begin to integrate IPs into their practices.

Correspondence: Donna M. D'Alessandro, MD, Department of Pediatrics, The University of Iowa Hospitals and Clinics, 200 Hawkins Dr, Iowa City, IA 52242-1009 (donna-dalessandro@uiowa.edu).

Accepted for publication March 5, 2004.

This study was supported by the Robert Wood Johnson Generalist Faculty Scholars Program, Princeton, NJ.

We thank Marcy Rosenbaum, PhD, for her help in survey form development.